June 10, 1996

Management and Prognosis of Life-threatening Bleeding During Warfarin Therapy

Author Affiliations

From the Division of General Medicine (Dr White and Ms McKittrick) and the Department of Medicine (Dr Takakuwa), University of California, Davis, Sacramento; the Division of Health Services Research (Mss Callahan and McDonell and Dr Fihn) and the Development Field Program (Dr Fihn), Veterans Administration Hospital, Seattle, Wash; and the National Consortium of Anticoagulation Clinics (Drs White, Takakuwa, and Fihn and Mss McKittrick, Callahan, and McDonell).

Arch Intern Med. 1996;156(11):1197-1201. doi:10.1001/archinte.1996.00440100095011

Background:  The incidence of explicity defined life-threatening bleeding during warfarin sodium therapy is largely unknown, as are the prognosis for and treatment of patients who have such bleeding. In addition, the location of the source of the life-threatening bleeding and the risk factors associated with life-threatening bleeding are not well-defined.

Objectives:  To determine the incidence of explicitly defined life-threatening bleeding during warfarin therapy, to identify the site of bleeding, to determine the risk factors for life-threatening bleeding, and to determine the risk of subsequent bleeding among patients receiving warfarin therapy.

Methods:  A cross-sectional prevalence study was conducted and data were combined with those obtained during prospective observation of a dynamic cohort of patients followed up in 2 university-affiliated and 3 Veterans Administration anticoagulation clinics.

Results:  For this study, 1999 patients were followed up for 3865 patient-years; 32 patients (11 women, 21 men, mean age of 60 years) met criteria for life-threatening bleeding, an incidence of 0.83 events/100 patient-years (95% confidence interval, 0.54-1.12). The most common indication for warfarin was to prevent thromboembolism because the patient had a mechanical heart valve (17/32 patients, 53%). The gastrointestinal tract was the definite or likely site of bleeding in 21 (66%) of the 32 patients. The prothrombin time ratio was longer than 2.0 or the international normalized ratio was longer than 4.5 in 16 (55%) of the 29 patients in whom a coagulation time was measured. Fourteen (44%) of the 32 patients had a history of peptic ulcer disease or gastrointestinal bleeding. Warfarin was restarted in 26 (81%) of the 32 patients. Twenty-five of 26 patients were followed up for a median of 30 months (range, 5-143 months); 14 (56%) of the 25 patients had a subsequent bleeding event, with 8 (57%) of the 14 having 1 or more additional life-threatening bleeding events (1 fatal) after a median of 11.5 months (range, 0.5-22 months).

Conclusions:  We conclude that in this cohort: (1) the incidence of life-threatening bleeding was rare, (2) the gastrointestinal tract was the site of bleeding in two thirds of the patients who experienced life-threatening bleeding, (3) most patients who experienced life-threatening bleeding had multiple risk factors for bleeding, including excessive anticoagulation, and (4) the risk of subsequent bleeding was extremely high among the patients who continued to receive warfarin therapy.(Arch Intern Med. 1996;156:1197-1201)